Accessing HD by bypassing BIOS?

[Here and Kickin']

My mom's Seagate ST38421A died after my sister's kids installed a game on
her computer (I do not think the kids caused the problem, however.) The
HD was factory installed into a Compaq 5441, and my attempt to diagnose
the problem is the first time that anyone has touched the innards of the
PC since it was assembled at the factory approximately six years ago.
The symptoms are a strange singing sound as the computer starts up (my mom
says that the sound was present for a time before the drive failed; the
sound is NOT the BIOS beeping). After trying to boot from the HD, the
CD-ROM, and the floopy disk. the computer announces a boot problem, and
asks me to provide a system disk.

The HD is the only device connected to the second IDE controller of the
computer. The motherboard is a Gigabyte GA-5SMM.

What I've done:

-Booted from drive A: system runs MS-DOS fine, but declares a "Invalid
drive specification" when I try to switch over to C:.

-Tried to use Compaq's restore disk, but the system could not find the HD;
the restore program asks me whether I want to restore again or exit the
program.

-Swapped the drive into my working Compaq 5204; no sound, but same error.
Same difficulties with accessing C:

-Tried to see whether the BIOS in my mom's computer has a HD diagnostic
function; it doesn't. (Just a fuction that allows me to choose the boot
device order. Fooling around with that didn't change the error message.)
The BIOS is from Award; according to the chip's label, it is
"PCI/PNP"-aware, and has a sticker that says "1.2B".

-Spent two nights downloading Debian Linux, burned the image on a CD-R,
and tried to boot off the CD; the same boot error message popped up.

-Downloaded Seagate's OS-independent diagnostic tool, and it doen't work
because the BIOS apparently can't see the HD. The program can talk
to the HD controllers, however.

-Finally, I popped off the jumper on the drive that changed the connection
from "cable select" to "slave". Booted from drive A:. The HD still could
not be found.

I cannot believe, aside from a possible mechanical problem that has caused
a major malfunction/destruction of the HD's inner workings, that it is
impossible to talk to the drive via software, somehow.

Does anyone have a suggested solution, or should I give up?

-d

 

[Ge]

After trying to boot from the HD, the CD-ROM, and the floopy disk. the
computer announces a boot problem, and asks me to provide a system disk.

Sounds like at least the boot partition of the HD is bad. Maybe the boot
sector, too. Maybe even the HD...

The HD is the only device connected to the second IDE controller of the
computer. [...]
-Finally, I popped off the jumper on the drive that changed the connection
from "cable select" to "slave". Booted from drive A:. The HD still could
not be found.

If the HD is the only device, it should probably be either "cable select"
(if you have a cable that supports this) and be connected to the master
port, or it should be "master" or "single" (if that option is provided in
the jumper settings). In any case, /not/ "slave".

-Booted from drive A: system runs MS-DOS fine, but declares a "Invalid
drive specification" when I try to switch over to C:.

If the partition is toast, you won't be able to switch to it. You could try
to repartition the drive (fdisk).

-Tried to use Compaq's restore disk, but the system could not find the HD;
the restore program asks me whether I want to restore again or exit the
program.

This might work better after repartitioning the drive with another tool. I
don't like Compaq's software... It's so "either it works as planned by
Compaq or it doesn't, but if it doesn't, you're pretty much lost"

-Swapped the drive into my working Compaq 5204; no sound, but same error.
Same difficulties with accessing C:

Same as above: if the partition is bad, you won't be able to access it in
the normal way.

-Downloaded Seagate's OS-independent diagnostic tool, and it doen't work
because the BIOS apparently can't see the HD. The program can talk
to the HD controllers, however.

That's odd. But I'd still try to repartition the disk (delete all
partitions, and then create at least one new partition). MS-DOS fdisk can
do that (usually).

It still could be that the HD is bad. This is not unheard of... I've got a
few of those lying around here

Gerhard

 

[Iago]

My mom's Seagate ST38421A died after my sister's kids installed a game on
her computer (I do not think the kids caused the problem, however.) The
HD was factory installed into a Compaq 5441, and my attempt to diagnose
the problem is the first time that anyone has touched the innards of the
PC since it was assembled at the factory approximately six years ago.
The symptoms are a strange singing sound as the computer starts up (my mom
says that the sound was present for a time before the drive failed; the
sound is NOT the BIOS beeping). After trying to boot from the HD, the
CD-ROM, and the floopy disk. the computer announces a boot problem, and
asks me to provide a system disk.

My mom's Seagate ST38421A is the only device connected to the second IDE controller of the
computer. The motherboard is a Gigabyte GA-5SMM.

What I've done:

-Booted from drive A: system runs MS-DOS fine, but declares a "Invalid
drive specification" when I try to switch over to C:.

-Tried to use Compaq's restore disk, but the system could not find the HD;
the restore program asks me whether I want to restore again or exit the
program.

-Swapped the drive into my working Compaq 5204; no sound, but same error.
Same difficulties with accessing C:

Does anyone have a suggested solution, or should I give up?

If you can find another ST38421A swap the electronics of the two
drives. I never tried this but there is a chance your mom's data is
still intact, only not accessible.

 

[Folkert Rienstra]

Sounds like at least the boot partition of the HD is bad. Maybe the boot
sector, too. Maybe even the HD...

In other words, you haven't got a faintest clue.

The HD is the only device connected to the second IDE controller of the
computer. [...]
-Finally, I popped off the jumper on the drive that changed the connection
from "cable select" to "slave". Booted from drive A:. The HD still could
not be found.

If the HD is the only device, it should probably be either "cable select"
(if you have a cable that supports this) and be connected to the master
port, or it should be "master" or "single" (if that option is provided in
the jumper settings).

In any case, /not/ "slave".

Clueless. Device1 is fine.

If the partition is toast, you won't be able to switch to it.

The partition isn't toast, the drive is.

You could try to repartition the drive (fdisk).

Without the bios seeing it?
In other words, you haven't got the faintest clue.

This might work better after repartitioning the drive with another tool.
Nope.

I don't like Compaq's software... It's so "either it works as planned by
Compaq or it doesn't, but if it doesn't, you're pretty much lost"

Sounds like it may apply to your braincells too.

Same as above: if the partition is bad, you won't be able to access it in
the normal way.


That's odd.

No, it is not. If the bios can't see it, so can't the diagnostic.

But I'd still try to repartition the disk (delete all
partitions, and then create at least one new partition). MS-DOS fdisk can
do that (usually).

Not without the bios seeing it.

It still could be that the HD is bad.

No, really?

This is not unheard of... I've got a few of those lying around here

That's so nice to know.

 

[Odie Ferrous]

If you can find another ST38421A swap the electronics of the two
drives. I never tried this but there is a chance your mom's data is
still intact, only not accessible.

It sounds to me more like a r/w head problem.

Try an identical logic board by all means (with this particular drive
you don't necessarily risk damaging it further) but be prepared to pay
silly money to have it recovered. Not an easy recovery at all if it is
indeed the r/w heads.


Odie

 

[Rod Speed]

My mom's Seagate ST38421A died after my sister's kids installed a
game on her computer (I do not think the kids caused the problem,
however.) The HD was factory installed into a Compaq 5441, and my
attempt to diagnose the problem is the first time that anyone has
touched the innards of the PC since it was assembled at the factory
approximately six years ago.

The symptoms are a strange singing sound as the computer starts up (my mom
says that the sound was present for a time before the drive failed; the sound
is NOT the BIOS beeping).

That is likely the drive recalibrating when it cant read the platters.

After trying to boot from the HD, the CD-ROM, and the floopy disk. the
computer announces a boot problem, and asks me to provide a system disk.

The HD is the only device connected to the second IDE controller of
the computer. The motherboard is a Gigabyte GA-5SMM.

What I've done:

-Booted from drive A: system runs MS-DOS fine, but declares a "Invalid drive
specification" when I try to switch over to C:.

-Tried to use Compaq's restore disk, but the system could not find the HD; the
restore program asks me whether I want to restore again or exit the program.

-Swapped the drive into my working Compaq 5204; no sound, but same error. Same
difficulties with accessing C:

-Tried to see whether the BIOS in my mom's computer has a HD
diagnostic function; it doesn't. (Just a fuction that allows me to
choose the boot device order. Fooling around with that didn't change
the error message.) The BIOS is from Award; according to the chip's
label, it is "PCI/PNP"-aware, and has a sticker that says "1.2B".

-Spent two nights downloading Debian Linux, burned the image on a
CD-R, and tried to boot off the CD; the same boot error message
popped up.

-Downloaded Seagate's OS-independent diagnostic tool, and it doen't
work because the BIOS apparently can't see the HD. The program can
talk to the HD controllers, however.

Then the drive has died and since its dead your PC too, it isnt something
basic like the ribbon cable, power or even the motherboard controller.

-Finally, I popped off the jumper on the drive that changed the
connection from "cable select" to "slave". Booted from drive A:. The HD still
could not be found.

All that indicates that the drive had died,
particularly its invisible to the Seagate diagnostic.

I cannot believe, aside from a possible mechanical problem that has
caused a major malfunction/destruction of the HD's inner workings,
that it is impossible to talk to the drive via software, somehow.

If the logic card fails, nothing will be able to see it.

Does anyone have a suggested solution, or should I give up?

Since you're trying so hard, presumably the data hasnt been
backed up properly. It may be possible to get the data back
by swapping the logic card from a known good identical model.

 

[Ge]

Sounds like it may apply to your braincells too.

It seems you're a lot better at communicating with hard disks than with
humans... Everybody to his own, I say -- happy diskin'!

Gerhard

 

[Ge]

Device1 is fine.

I thought that the jumper setting would have to be "single" (if provided)
or "master" in case of a single drive on the bus. You seem to say that if
there is only one drive on the bus, it can be any setting ("single",
"master" or "slave").

You also seem to imply that it is ok to connect a drive set to "cable
select" to the slave connector on a cable with connectors that provide the
cable select signal.

Is this true? If it is, why do some disk manufacturers provide a "single"
position for the jumpers? Why do they write in their disk manuals (at least
a few that I've seen) that single drives should be set to "master" (or
"single" if provided)?

Gerhard

 

[Folkert Rienstra]

I thought that the jumper setting would have to be "single" (if provided)
or "master" in case of a single drive on the bus.

Nope. Nor for a long time now
And 'single' is only a setting used by one particular drive maker.

You seem to say that if there is only one drive on the bus, it can be any
setting ("single", "master" or "slave").

It can be device0 or it can be device1. How that translates in jumper settings
can be a different matter.

You also seem to imply that it is ok to connect a drive set to "cable select"
to the slave connector on a cable with connectors that provide the cable
select signal.

No, I don't. And cable select is not a signal, it is a key.

Is this true?

No, that is a NoNo. That makes for an open ended cable.

If it is, why do some disk manufacturers provide a "single"
position for the jumpers?

So they don't wait forever for a nonexistent slave to report OK at
power-up selftest.

Why do they write in their disk manuals (at least a few that I've seen)
that single drives should be set to "master" (or "single" if provided)?

To be compatible with old MoBos that didn't support booting from
other drives than the master on IDE0.

 

[Rod Speed]

Folkert Rienstra wrote

I thought that the jumper setting would have to be "single"
(if provided) or "master" in case of a single drive on the bus.
You seem to say that if there is only one drive on the bus,
it can be any setting ("single", "master" or "slave").

Many drives will work fine jumpered as slave as the only drive on a cable.

You also seem to imply that it is ok to connect a drive
set to "cable select" to the slave connector on a cable
with connectors that provide the cable select signal.

Is this true?

Yes, many drives will work like that.

Its not ideal with modern high speed drives tho because
you have the unterminated 'master' connector going
nowhere and you can get reflections off that cable stub.

If it is, why do some disk manufacturers
provide a "single" position for the jumpers?

Basically bad design.

Why do they write in their disk manuals (at least a few that I've seen)
that single drives should be set to "master" (or "single" if provided)?

Basically thats the historical use of the master/slave setting. That
doesnt mean that the drive wont work jumpered as slave when its
the only drive on a cable tho, plenty will work fine like that.

 

[Joep]

It seems you're a lot better at communicating with hard disks than with
humans...

It's not that, he has a small dick ans he is a big dick, that's why he acts
as he does.

 

[Joep]

It's not that, he has a small dick ans he is a big dick, that's why he acts
as he does.

right twice in one thread

 

[chrisv]

right twice in one thread

Nice mangling.

 

[Ge]

Many drives will work fine jumpered as slave as the only drive on a cable.

Thanks for all the clarifications. It seems it's still safer to jumper a
single drive as master (even though the consensus is that most drives work
well in this situation).

Gerhard

 

[Rod Speed]

Rod Speed wrote

Thanks for all the clarifications. It seems it's still safer
to jumper a single drive as master (even though the
consensus is that most drives work well in this situation).

Sure, or use cable select.

 

[Robert J. Stevens]

Sounds like at least the boot partition of the HD is bad. Maybe the boot
sector, too. Maybe even the HD...

I have a IBM Travelstar that quit. Drive clicks when try to access. I was able
to see it once on the Laptop when I used HDDREG 1.41. It showed that the block
was bad and sat there for awhile. I should have tried to start at 1M but didnt.
Is there a program out there that will rewrite the MBR [track zero] without
rying to reasd first. Other Prgrams I have sych as MBRWORK says read failed. I
am trying to look at the drive with KNOPPIX right now.
TIA
Bob in Wisconsin